We analyze samples of nearby clusters taken from the Abell catalog and the X-ray Sample of Bright Clusters ( De Grandi et al 1999 ) including a wide range of X-ray luminosities .
Using the usually adopted background subtraction procedures , we find that galaxies in clusters selected by means of their X-ray emission show a flat luminosity function ( faint end slope \alpha \simeq - 1.1 ) consistent with that derived for galaxies in the field and groups .
By contrast , the sample of Abell clusters that do not have an X-ray counterpart shows a galaxy luminosity function with a steep faint end ( \alpha \simeq - 1.6 ) .

We investigate the possibility that cD halos could be formed by the disruption of galaxies in rich relaxed clusters that show an apparently flat faint end galaxy luminosity function ( Lopez-Cruz et al 1997 ) .
We find that clusters dominated by a central cD galaxy ( Bautz-Morgan classes I and II ) show the same systematic trend : X-ray selected clusters have flatter faint end slopes than those clusters with no detected X-ray emission .
Thus , it is likely the X-ray selection and not the cluster domination by central galaxies what correlates with background decontamination estimates of the galaxy luminosity function .
Moreover , no significant correlation between X-ray luminosity and the galaxy LF faint end slope is found .
These results do not support a scenario where flat faint end slopes are a consequence of cD formation via the disruption of faint galaxies .
We argue that the clusters without X-ray emission are strongly affected by projection effects which give rise to spurious faint end slopes estimated using background subtraction procedures ( Valotto et al 2001 ) .